Synthesis and structural and magnetic characterization of the alkoxo

University of North Carolina, Chapel Hill, North Carolina 27514. Synthesis and Structural ... magnetic properties of chromium(II1) dimers of these typ...
1 downloads 0 Views 517KB Size
Inorganic Chemistry, Vol. 15, No. 5, 1976 1179

Alkoxo-Bridged Cr"' Dimer

Contribution from the Department of Chemistry, University of North Carolina, Chapel Hill, North Carolina 27514

Synthesis and Structural and Magnetic Characterization of the Alkoxo-Bridged Chromium(II1) Dimer Di-p-methoxo-bis[bis(3-chloro-2,4-pentanedionato)chromium(III)] EVA DIXON ESTES, RAYMOND P. SCARINGE, WILLIAM E. HATFIELD, and DEREK J. HODGSON'

Received October 31, 1975

AIC50784J

The complex di-~-methox~bis[bis(3-chloro-2,4-pentan~onato)chromiu(III)] , [Cr(3-Cl(acac))20CH3]2, has been synthesized and its crystal structure has been determined from three-dimensional counter x-ray data. The material crystallizes in the monoclinic space group P21/c with two dimeric formula units in a cell of dimensions a = 7.794 (6) A, b = 18.189 (18) A, c = 11.232 (8) A, p = 112.82 (5)'. Least-squares refinement of 3092 independent data has led to a final R factor (on F)of 0.045. The complex consists of two chromium(II1) ions which are symmetrically bridged by two methoxide groups, the remaining coordination sites being occupied by the oxygen atoms of two bidentate 3-Cl(acac) ligands. The coordination around each chromium center is roughly octahedral; the average Cr-O(1igand) bond length is 1.957 (5) 8,while the Cr-Cr separation is 3.025 (2) A and the C r - O C r bridging angle is 101.07 (9)O. The 3-Cl(acac) ligands are approximately planar. The magnetic susceptibility of a powdered sample of the complex has been examined in the temperature range 4-78 K. The dimer exhibits an antiferromagnetic exchange interaction, with the best fit to the Van Vleck equation including biquadratic exchange yielding 2 J = -7.34 cm-' and j = +0.38 cm-l with (g) = 1.901. These data lead to a singlet-triplet splitting of -9.82 cm-'.

Introduction The study of the magnetic interactions in hydroxo- and alkoxo-bridged dimers of copper(I1) and chromium(II1) has received extensive recent It has been d e m ~ n s t r a t e d ~for - ~ the simplest case, i.e. the dihydroxobridged dimers of copper(II), that the magnitude and sign of the exchange integral J are principally determined by the value of the Cu-0-Cu bridging angle 4, although other authors7?* have noted that distortions from planar geometry at,the metal center also influence the magnitude of J. For the analogous chromium(II1) complexes the magnetic exchange is greatly complicated by the presence of more than one unpaired electron at each metal center, which leads to unpaired spin in the ?r system of the dimer in addition to that in the Q system. There is, however, one simplifying feature in that the geometry at chromium in all complexes of the types [CrL20H]2"+ and [CrL20R]2"+ (where L is a bidentate ligand) is pseudooctahedral and, therefore, approximately constant throughout the series. Hence, the probable influence7 of changes in the geometry at the metal on the magnetic exchange does not intrude here. We have, therefore, initiated studies on the structural and magnetic properties of chromium(II1) dimers of these types and have reported e l ~ e w h e r e ~ *the ~ -results '~ of our studies of several hydroxo-bridged species. We here report the synthesis, structure, and magnetic properties of the first completely characterized dialkoxo-bridged dimer of chromium, the complex being di-1.~-methoxo-bis[bis(3-chloro-2,4-

pentanedionato)chromium(III)]

.

Experimental Section Synthesis. A saturated solution of the ~ o m p l e x tris(3-chloro'~ 2,4-pentanedionato)chromium(III) in methanol was refluxed for several days. The fine green precipitate which formed was removed by filtration and recrystallized twice from benzene-heptane solutions, yielding the green desired complex, diy-methoxo-bis[bis(3-chloro2,4-pentanedionato)chromium(III)]. This procedure is entirely analogous to that used by Kasuga et aI.l4 for the synthesis of the 3-bromo analogue. Anal. Calcd for C22H30C14Cr2010: C, 37.7 1; H, 4.29; C1, 20.29. Found: C, 37.96; H, 4.12; C1, 20.46. Single crystals for the crystallographic study (vide infra) were obtained by recrystallization from dichloromethane to which a few drops of ethanol had been added. Magnetic Swceptibility. The magnetic susceptibilityof a powdered sample of the dimer was measured from 4 to 78 K using a Foner type PAR vibrating-samplemagnet~meter.'~ All measurements were made at a field strength of 10000 G. Temperatures were measured by a

calibrated gallium arsenide diode, and mercury tetrathiocyanatocobaltate(I1) was used as a susceptibility standard.16 Collection and Reduction of the X-Ray Data. On the basis of Weissenberg and precession photographs, the crystals were assigned to the monoclinic system. The observed systematic absences are h01 for I odd and OkO for k odd, which suggests that the space group is P21/c (C2h5). The cell constants, obtained by least-squares method, are a = 7.794 (6), A, 6 = 18.189 (18) A, c = 11.232 (8) A, and p = 112.82 ( 5 ) O . The observations were made at 20.5 OC with the wavelength assumed as X(Mo Kal) = 0.7093 A. A density of 1.585 g cm-3 calculated for two dimeric formula units in the cell compares well with a value of 1.57 (3) g cm-3 observed by flotation in aqueous zinc bromide solution. Hence, in space P21/c, the dimer is constrained to lie on a crystallographic inversion center. Diffraction data were collected from an irregular shaped crystal and (Oil). with faces (loo), (iOO), (OlO), (Oio), (01 l), ( O i l ) , (01 I), The separations between opposite pails of faces were as follows: (100) and (loo), 0.0_92 cm; (?lo) and (OlO), 0.030 cm; (01 1) and (01 l ) , 0.041 cm; (01 1) and (01 l), 0.002 cm. The crystal was mounted on a glass fiber normal to the (100) planes, and in this orientation intensity data were collected on Picker four-circle automatic diffractometer equipped with a graphite monochromator and using Mo K a radiation. The mosaicity of the crystal was examined by means of the narrow-source, open counter w-scan technique" and was judged to be acceptable. Twelve reflections, accurately centered through a narrow vertical slit at a takeoff angle of l.Oo, formed the basis for the least-squares refinement of cell parameters and orientation using the logic documented by Busing and Levy1* for the PDP-B/L computer. Intensity data were collected at a takeoff angle of 1.3'; at this angle the intensity of a typical reflection was approximately 85% of maximum as a function of takeoff angle. The receiving aperture was 5.0 mm high by 5.0 mm wide and was positioned 32.0 cm from the crystal. The data were collected by the 8-28 scan technique at a scan rate of l.OO/min. Allowance was made for the presence of both Kal and Ka2 radiations, the peaks being scanned from -0.75' in 20 below the calculated Kal peak position to +0.75O in 28 above the calculated Ka2 position. Stationary-counter,stationary-crystalbackground counts of 10 s were taken at each end of the scan. The pulse height analyzer was set for approximately a 90% window, centered on the Mo K a peak. A unique data set having 2' I 28 I 58' was gathered; a total of 3980 intensities were recorded. The intensities of three standard reflections, measured after every 100 reflections, declined steadily throughout the run so that the average intensity had dropped by approximately 8% by the end of the data collection. This decline in intensity as a function of exposure time was allowed for in the structure refinement. l9 Data processing was carried out as described by Corfield et aL20 After correction for background, the intensities was assigned standard

1180 Inorganic Chemistry, Vol. 15, No. 5, 1976

Hodgson et al.

Table I. Positional Parameters (X l o 4 ) for [Cr(3-Cl(acac)),0CH31, Atom Cr c11 c12 01 02 03 04 OB C1A C2A C3A C4A C5A C1B C2B C3B C4B C5 B CME HIAa H2A H3A H4A H5 A H6A H1B H2B H3B H4B H5B H6 B HM 1 HM2 HM3

a

Y

Z

224 (0) 2498 (1) -1200 (1) 1137 (1) 554 (1) -87 (1) -290 (1) -623 (1) 2323 (2) 1690 (1) 1713 (1) 1150 (1) 1220 (2) -539 (2) -436 (1) -725 (2) -646 (1) -994 (2) -1361 (2) 274 (4) 235 (4) 230 (4) 167 (5) 75 (3) 120 (3) -28 (2) 398 (3) 465 (5) 75 (3) -146 (4) 407 (3) 362 (3) 169 (2) 348 (3)

1390 (0) 2367 (1) 5071 (1) 2336 (2) 1024 (2) 1867 (2) 2922 (2) 242 (2) 3200 (5) 2481 (2) 2017 (3) 1319 (2) 873 (4) 2930 (4) 2839 (2) 3785 (2) 3778 (2) 4813 (3) 629 (3) 274 (5) 406 (8) 315 (6) 1(8) 50 (4) 168 (5) 208 (4) 173 (4) 156 (8) 442 (6) 482 (5) 45 (6) 384 (5) 6 (4) 435 (6)

X

1067 (0) 5108 (1) 465 (2) 1212 (3) 3209 (2) -984 (2) 2784 (2) 713 (2) 1901 (8) 2280 (4) 3702 (4) 4121 (3) 5701 (4) -2820 (5) -967 (4) 680 (4) 2454 (3) 4130 (5) 839 (5) 179 (8) 239 (9) 80 (10) -468 (13) 589 (6) 695 (8) -376 (6) 304 (7) 312 (12) -427 (8) 393 (7) 474 (8) -184 (8) -79 (6) 40 (8) Hydrogen atom parameters are

Figure 1. View of the dimeric [Cr(3-Cl(acac)),OCH,], molecule. Hydrogen atoms have been omitted for clarity. Unlabeled atoms are related to labeled atoms by the inversion center.

X103.

deviations according to the formula2’

41)= [C + 0.25(ts/tb)*(BH+ BL)+ and the value of p was selected as 0.050. The values of I and o(1) were corrected for Lorentz-polarization effects and for absorption factors. The absorption coefficent for this compound for Mo K a radiation is 11.85 cm-’, and for the crystal chosen the transmission coefficientsZ2evaluated by numerical integration ranged from 0.637 to 0.741. Of the 3980 data collected, 3092 were greater than 3 times their estimated standard deviations; only these data were used in the subsequent structure analysis and refinement. Solution and Refinement of the Structure. The position of the one independent chromium atom was determined from a three-dimensional Patterson function, and two cycles of least-squares refinement of this position were run. All least-squares refinements in this analysis were carried out on F, the function minimized being Cw(lFol - lFc1)2and the weights w being taken as 4FO2/o2(Fo2).In calculations of Fc, the atomic scattering factors for Cr and CI were from Cromer and , ~ ~those for H from Waber,23 those for 0 and C from I b e r ~ and Stewart, Davidson, and S i m p ~ o n .The ~ ~ effects of the anomalous dispersion of Cr and CI were included in the calculation of F,, the values of Af’ and Af” being taken from Cromer and Liberman.26 The remaining nonhydrogen atoms were located in subsequent difference Fourier maps. Isotropic least-squares refinement of these 19 atoms led to values of the conventional agreement factors

of 0.139 and 0.228, respectively. Anisotropic refinement of these atoms gave R1 = 0.073 and R2 = 0.125. The positions of 14 of the 15 hydrogen atoms were located in subsequent difference Fourier syntheses. The remaining hydrogen atom was placed in the position calculated on the basis of tetrahedral geometry at the carbon atom with a C-H distance 0.95 A,27and a least-squares calculation in which

Figure 2. Coordination around the chromium centers in [Cr(3Cl(acac)),OCH,],, all hydrogen atoms were refined isotropically while nonhydrogen atoms were refined anisotropically gave values of R1 and R2 of 0.045 and 0.058, respectively. The value of R2 showed some dependence on sin 0, and so the weighting scheme was modified by dividing the weights w for reflections with 2’ I26 I20’ by 1.80. In most cases in which we have used our graphite monochromator a similar correction has been necessary. The final values of Fo and Fc suggested that no correction for secondary extinction was necessary. In the final cycle of least-squares refinement no parameter underwent a shift in excess of 1OOh of its estimated standard deviation, which is taken as evidence that the refinement had converged. A final difference Fourier synthesis was featureless. The final values of R1 and R2 were 0.045 and 0.056, respectively. The positional and thermal parameters derived from the final least-squares cycle are given in Tables I and 11. A compilation of observed and calculated structure amplitudes is available.28

Description of the Structure The structure consists of dimeric [Cr(3-Cl(acac))20CH3]2 (3-Cl(acac) = 3-chloro-2,4-pentanedione)units which are well-separated from one another. The geometry of the dimer is shown in Figure 1, and the inner coordination sphere around the chromium atoms is shown in Figure 2. The geometry around the chromium atoms is roughly ligands and octahedral with two cis 3-chloro-2,4-pentanedione two cis methoxy groups occupying the coordination sites. The bond lengths and angles in the dimer are given in Tables 111 and IV. The Cr-0-Cr-0 bridging unit is strictly planar, there being a crystallographic inversion center in the middle of the dimer. The methoxy carbon atom CME is considerably out of this plane, its deviation from the plane being 0.58 A. For trigonal geometry at oxygen, this value would be zero, while for tetrahedral geometry it would be 1.15 A. Hence,

Inorganic Chemistry, Vol. 15, No. 5, 1976 1181

Alkoxo-Bridged Cr"* Dimer Table 11. Thermal Parameters (A') for [Cr(3Cl(acac)),OCH,]," Atom Cr c11 c12 01 02 03 04 OB C1A C2A C3A C4A C5 A C1B C2B C3B C4B C5 B CME

u 1 1

U',

u33

229 (2) 720 (5) 788 (6) 420 (9) 283 (8) 278 (8) 263 (7) 290 (7) 909 (30) 446 (13) 377 (12) 235 (10) 359 (14) 449 (16) 369 (12) 457 (13) 402 (12) 466 (16) 624 (19)

288 (2) 470 (4) 1227 (9) 372 (9) 396 (9) 413 (9) 483 (10) 248 (7) 435 (17) 342 (12) 320 (12) 378 (12) 572 (19) 628 (20) 348 (11) 491 (14) 393 (12) 648 (20) 293 (12)

322 (2) 1262 (9) 691 (5) 433 (9) 430 (9) 430 (9) 343 (8) 349 (7) 774 (25) 365 (12) 526 (14) 420 (12) 817 (22) 774 (22) 446 (12) 395 (12) 313 (10) 406 (14) 483 (15)

Vl 2 5 (1) -282 (4) 24 (6) -53 (7) -34 (7) 16 (7) 20 (7) 40 ( 5 ) -70 (18) 1(10) -67 (10) -25 (9) -7 (13) -16 (14) -12 (9) -17 (12) 38 (10) 123 (141 92 (12)

'13

'23

116 (1) 401 (6) 410 (5) 210 (8) 170 (7) 180 (7) 100 (6) 113 (6) 446 (23) 109 (10) 109 (1 1) 87 (9) 318 (15) 420 (16) 247 (10) 224 (1 1) 125 (9) 108 (12) 155 (14)

48 (1) -119 (5) 496 (6) -45 (7) 23 (7) 84 (7) 98 (7) 61 (6) -154 (18) 24 (9) 52 (11) 119 (10) 159 (17) 111 (17) 12 (10) 126 (11) 49 (9) 188 (14) 133 (11)

Atom

U

Atom

U

Atom

U

H1A H2A H3A H4A H5 A

11 (2) 14 (3) 13 (2) 19 (3) 9 (1)

H6A H1B H2B H3B H4B

10 (1) 7 (1) 8 (1) 17 (31 11 (2)

H5B H6B HM1 HM2 HM3

10 (2) 11 (2) 10 (2) 8 (1) 11 (2)

a Anisotropic parameters are X104; isotropic parameters, XlO'. The form of the anisotropic thermal ellipsoid is exp[-2nz(U1,hza*' U2zk2b*2+ U3311c*zt 2U12hka*b* t 2U,,hla*c* + 2UZ3klb*c*)].

Table 111. Interatomic Distances (A) in [Cr(3Cl(acac)),0CH3], Cr-Cr' Cr-OB Cr-OB' 0-01 Cr-02 Cr-03 Cr-04 01C2A

3.025 (2) 1.958 (2) 1.960 (2) 1.952 (2) 1.962 (2) 1.956 (2) 1.959 (2) 1.275 (3)

02C4A 03C2B 04C4B ClAC2A C2AC3A C3AC4A C4A45A

1.269 (3) 1.258 (3) 1.265 (3) 1.500 (5) 1.396 (4) 1.403 (4) 1.504 (4)

ClBC2B C2BC3B C3BC4B C4BC5B C3AC11 C3BC12 OBCME

1.499 (4) 1.412 (4) 1.393 (4) 1.510 (4) 1.749 (3) 1.747 (3) 1.403 (3)

in this complex the hybridization at the bridge cannot be described as either sp2 or sp3. The two independent bridging Cr-0 bond distances are 1.960 ( 2 ) and 1.958 ( 2 ) A which shows that the oxygen atoms are symmetrically bound to the two Cr atoms. These values are longer than in [Cr(phen)20H]f34H20 and [Cr(phen)20H]2Clp6H20 and similar to those in [Cr(gly)20H]2. The Cr-Cr separation of 3.025 A is longer than those found in these dihydroxobridged chromium dimers; however, the bridging Cr-0-Cr angle of 101.07 (9)O is intermediate between the values of 98.2 ( 2 ) , 102.1 (3), and 102.7 ( 5 ) O found for the glycinato1° and the iodide12 and chloride" salts of the phenanthroline complex, respectively. Although the actual geometry of the dimer is Ci(i), it approximates C2h ( 2 / m ) , rather than the apparently equally possible D2 (222). In this respect it is similar to all related hydroxo-bridged dimers except the phenanthroline complexes; in these latter complexes, as has been noted elsewhere,2 the C2h conformer is unobtainable because of interligand H-H contacts. The 3-chloro-2,4-pentanedionatoligands are nearly planar, with the greatest deviation of any atom from the best least-squares plane through the ligand atoms and the chromium atom being 0.04 A for both the A ring and the B ring. The geometry of the ligand is similar to that found for the parent 2,4-pentanedionato ligand in variety of metal comp l e ~ e s . ~ ~The - ~ 'average 0-0 "bite" of 2.736 A found here is slightly smaller than the values of 2.751 and 2.786 A found in the m a n g a n e ~ e ( I I 1 )and ~ ~ chromium(III)30 complexes M(acac)3. The average Cr-0 bond length of 1.957 (4) A is not significantly longer than the values of 1.952 (8) and 1.952

+

Table IV. Interatomic Angles (deg) in [Cr(3Cl(acac)),0CH3], Cr-OBCr' Cr-OBCME Cr'-OBCME OBCr-01 OBCr-02 OBCr-03 OBCr-04 OB-Cr-OB' 01Cr-02 01C r - 0 3 01Cr-04 0141-OB' 02Cr-03 02-0-04 0243-OB' 0341-04 03Cr-OB' 04-0-OB' Cr-OlC2A OlC2AClA

101.07 (9) 125.0 (2) 125.7 (2) 172.6 (1) 90.94 (8) 91.84 (8) 94.4 (1) 78.93 (9) 88.81 (8) 88.72 (9) 93.0 (1) 93.68 (9) 176.4 (2) 89.05 (9) 92.37 (9) 88.52 (9) 90.35 (9) 173.2 (1) 129.7 (2) 114.8 (3)

0142A-C3A ClAC2AC3A C2AC3AC4A C2AC3ACll Cll43A44A C3AC4A-02 C3AC4AC5A C5AC4A-02 C4A-02-0 Cr-0342B 03C2BClB 0342BC3B ClBC2BC3B C2BC3BC4B C2BC3BC12 C1243B44B C3BC4B-04 C3B44BC5B C5BC4B-04 C4B-04Cr

123.3 (2) 121.9 (3) 125.3 (2) 117.3 (2) 117.4 (2) 122.8 (2) 121.4 (2) 115.7 (3) 130.0 (2) 130.4 (2) 115.8 (2) 122.8 (2) 121.4 (2) 124.9 (2) 117.1 (2) 118.0 (2) 123.5 (2) 121.1 (2) 115.4 (2) 129.8 (2)

(10) A found in chromium(II1) complexes of 2,4p e n t a n e d i ~ n e . ~In~ the ? ~ ~complex Cr(3-Cl(acac))3, which is the precursor to the dimeric material studied here, however, it is apparent that the Cr-0 bonds are weaker than in Cr(acac)3 since, in the former, it is possible to remove one ligand in alcoholic medium to form the present dimer whereas Cr(acac)3 is stable to refluxing in alcoh01.l~ The chelating 0-0-0 angles of 88.5 and 88.7' are acute, whereas in rep0rted~~9~l chromium(II1) complexes of acac they are obtuse. Discussion The plot of molar magnetic susceptibility as a function of temperature exhibits a maximum at about 20 K, indicating antiferromagnetic coupling. The data were fitted to the expression Xm

N2P2 =F

2e(2J-6.5j)/kT+ 1oe(6J-13.5j)/kT + 2ge(12J-9j)/kT 1 + 3e(W4.5i)/kT+ se(6J-13.5j)/kT + 7e(1ZJ-9j)/kT

1182 Inorganic Chemistry, Vol. 15, No. 5, 1976

Hodgson et al. Supplementary Material Available: Listing of structure factor amplitudes (18 pages). Ordering information is given on any current masthead page.

References and Notes

x

t r -

(1) D. J. Hodgson,Prog. Inorg. Chem., 19,173 (1975), and references therein. (2) D. J. Hodgson in “Extended Interactions between Metal Ions”, L. V. Interrante, Ed., American Chemical Society, Washington, D.C., 1974, Chapter 9. (3) D. L. Lewis, W. E. Hatfield, and D. J. Hodgson, Inorg. Chem., 11,2216 (1972). (4) K. T. McGregor, N. T. Watkins, D. L. Lewis, R. F. Drake, D. J. Hodgson, and W. E. Hatfield, Inorg. Nucl. Chem. Lett., 9, 423 (1973). (5) D. L. Lewis, K. T. McGregor, W. E. Hatfield, and D. J. Hodgson, Inorg. Chem., 13, 1013 (1974). (6) E. D. Estes, W. E. Hatfield, and D. J. Hodgson, Inorg. Chem., 13, 1654

b

r 1 b

(1974).

0 , 0

40

BO

T i’K1

Figure 3. Temperature dependence of the magnetic susceptibility of [Cr(3-Cl(acac)),0CH31,. The experimental data are shown as squares. The dashed line represents the best fit to the dimer (S = 3/2, 3 / 1 2 ) equation in the absence of biquadratic exchange. The solid line represents the best fit to the dimer equation including biquadratic exchange.

where J and j are defined by the exchange Hamiltonian

(7) E. Sinn, paper presented at the American CrystallographicAssociation Meeting, Charlottesville, Va., March 1975. (8) M. D. Glick and R. L. Lintvedt, paper presented at the 167th National

Meeting of the American Chemical Society, Los Angeles, Calif., April

1974. (9) D. J. Hodgson, J. T. Veal, W. E. Hatfield, D. Y. Jeter, and J. C. Hempel, J . Coord. Chem., 2, 1 (1972). (1 0) J. T. Veal, W. E. Hatfield, D. Y. Jeter, J. C. Hempel, and D. J. Hodgson, Inorg. Chem., 12, 342 (1973). (1 1) J. T. Veal, W. E. Hatfield, and D. J. Hodgson, Acta Crystallogr., Sect. B, 29, 12 (1973). (12) R. P. Scaringe, P. Singh, R. P. Eckberg, W. E. Hatfield, and D. J. Hodgson, Inorg. Chem., 14, 1127 (1975). (13) J. P. Collman, R. A. Moss, H. Maltz, and C. C. Heindel, J . Am. Chem. SOC.,83, 531 (1961). (14) K. Kasaga, T. Itou, and Y. Yamamoto, Bull. Chem. Sot. Jpn. 47, 1026

. .,.

(,*,1974)

S. Foner, Rev. Sci. Instrum., 30, 548 (1959). B. N. Figgis and R. S. Nyholm, J . Chem. SOC.,4190 (1958). T. C. Furnas, “Single Crystal Orienter Instruction Manual”, General Electric Co., Milwaukee, Wis., 1957. W. R. Busing and H. A. Levy, Acta Crystallogr., 22, 457 (1967). E. D. Estes and D. J. Hodgson, Inorg. Chem., 14, 334 (1975). P. W. R. Corfield, R. J. Doedens, and J. A. Ibers, Inorg. Chem., 6, 197

The data were fitted using a general nonlinear least-squares program written in our laboratory, the function minimized being F = C.w.( 1 1 Xl.Obsd - ~ i ~with~ thel weights ~ ~ being ) as~ signed wi = 1/xiobsd. The best fit to the data, holding j = 0 and letting g and J vary, is displayed in Figure 3, where the (1967). final values were g = 2.05 (1) and J = -5.04 (6) cm-’. The W. R. Busing and H. A. Levy, J . Chem. Phys., 26, 563 (1957). fit is at best unimpressive but this is not surprising since it has For a descriptionof the programs used, see E. D. Estes and D. J. Hodgson, Inorg. Chem., 12, 2932 (1973). often been found necessary to include a term biquadratic in T. Cromer and J. T. Waber, Acta Crystallogr., 18, 104 (1965). spin to obtain good fits for dimeric Cr(II1) c o m p l e x e ~ . ~ ? ~ ~ , ~ ~D. J. A. Ibers, “International Tables for X-Ray Crystallography”, Vol. 11, The best fit, with the inclusion of the biquadratic exchange Kynoch Press, Birmingham, England, Talbe 3.3 1A. parameter, j , yielded g = 1.901 (4),J = -3.67 (3) cm-’, and R. F. Stewart, E. R. Davidson, and W. T. Simpson, J. Chem. Phys., 42, 3175 (1965). j = 0.381 (4)cm-l, where the quantities in parentheses are D. T. Cromer and D. Liberrnan, J . Chem Phys., 53, 1891 (1970). the estimated standard deviations of the respective parameters M. R. Churchill, Inorg. Chem., 12, 1213 (1973). from the last cycle of least-squares for which there were 50 Supplementary material. B. Morosin and J. B. Brathode, Acta Crystallogr., 17, 705 (1964). observations and three variables. These data led to a sinB. Morosin, Acta Crystallogr., 19, 131 (1965). glet-triplet splitting of -9.82 cm-’. The theoretical suscepC. E. Wilkes and R. A. Jacobson, Inorg. Chem., 4, 99 (1965). tibilities using these values for g, J, and j are also presented H. Ikeda, I. Kimura, and N. Uryu, J . Chem. Phys., 48,4800 (1968). J. T. Veal, D. Y. Jeter, J. C. Hempel, R. P. Eckberg. W. E. Hatfield, in Figure 3, and it can be seen that the comparison with and D. J. Hodgson, Inorg. Chem., 12, 2928 (1973). experiment is greatly improved. The positive sign obtained (34) C. Kittel, Phys. Rev. 120, 35 (1960). for j in this study is significant in that it supports Kittel’s (35) J. C. M. Henning, J. H. den Boef, and G. G. P. van Gorkom, Phys. Rea B, 7,1825 (1973). It should be noted that these authors obtained a negative exchange striction mechanism.34 This point is discussed in value of j because they wrote the Hamiltonian X = JSlaS2 + j ( S 1 6 2 ) ~ . some detail elsewhere.35 Their result is, therefore,equivalent to a positivej using our Hamiltonian Registry No. [Cr(3-Cl(acac))20CH3]2, 585 12-49-5. (eq 2).